T&abdron Letters No.46, pp. 4629-4634, 1967. Pergamon press Ltd. Rinted in Great Britain.

CTCLCPROPTLCARBIRTL FREE RADICALS, BBXACTCLOPROPTLBTRANB

J. C. Martin, John E. Schultz, and Jack W. Timberlake

Department of Chemistry, University of Illinois

Urbana, Illinois 61801

(Received in USA 10 July 1967)

Several cases have been reported (I) in which radical-forming reactions are accelerated by

the introduction of a cyclopropyl substituent directly on the incipient radical center. It is

not clear, however, just how the acceleration is best explained. A concerted multiple bond

cleavage reaction leading directly to the allylcsrbinyl system might reflect a driving force

from relief of strain accompanying the opening of a three-membered ring. Transition state charge

polarization (2) in a radical-forming reaction could produce an acceleration reflecting a re-

semblance of the transition state to the corresponding cyclopropyl carbonium ion. A very small.

fraction of the large accelerating effect of s cyclopropyl substituent seen in reactions leading

to carbonium ions would be sufficient to explain the rate data available for reactions leading

to cyclopropylcarbinyl radicals (3). This paper reports kinetic evidence for a series of de-

composition reactions in which acceleration associated with cyclopropyl substituents is not

related to cleavage of the cyclopropyl rings and in which transition

effects seem to be unimportant.

Overherger and Berenbaum (1 g,h) report the decomposition (80°,

state charge polarization

toluene) of 2-2’-azobis-

2-cyclopropylpropionitrile (Ib) to be twenty times faster than that of azobisisobutyronitrile

(AIBN, In). Since the only product with cyclopropyl rings intact (III) is formed in only 19%

yield in solution decomposition of Ib (A), the possibility remains that all* or most of the

observed acceleration could be attributed to a concerted cleavage of C-N and C-C bonds. In

*A three-bond cleavage (similar to that illustrated by V) propane ring, would yield one radical with the cyclopropa:e

involving only a single cyclo- ring intact.

4629

46X Co.46

certain other cases acceleration in the rate of formation of cyclopropylcarbinyl radicals is

thought to be the result of “manifestations of the ring strain only” (5).

Ib

TABLE 1

Decomposition of Azo Compounds

f‘l r1 R2 - f:

-N=B-C-R I 2

R3 R3

In Toluene at 80.2'

k,Se? relative

Compound Rl R2 R3 x lo4 rates AH* As* AG*

Ia rn3 cH3 CN 1.4sa 1.0 30.7 ;t O.la 10.4 + 0.2' 27 .o

Ib cH3 cycle-C3H5 CN 37.sb 25.8 27.7 2 0.4b 8.4 + L2b 24.7

Ic _eyclo-C3H5 cycle-C3H5 CN 347c 240 23.7 ;f 0.7' 1.5 f 2.5' 23.2

In Diphenylether at 135.0'

IIa CH3 CH3 lx3 0.0094 1.0 42.2 t_ 0.3' 16.2 + O.fid 35.6 fib cycle-C3H5 m3 cH3 0.252 26.8 37.8 2 0,3e 12.4 + 0,6e 32.7

IIC cyclo-c3115 cp&-C3H5 CH3 3.41 362 35.6 i 0.2f 12.1 + 0,6f 30.6 XII cycle-C3H5 cycle-C3H5 c&lo-C3H5 23.9 2,540 34.3 + 0.38 L2.8 + 0.88 29.1

IIe ~J~&-C~H~ cycLo-C3H5 A-C3H7 2.70 286 38.0 + 0-P 17.6 + 0.6h 30.8

aCalculated from data by Y. P. Van Hook and A. V. Tobolsky, 2. &I. C&em. SQC., 80, 779 (1958). _--

bCalculated from data in reference lh. ‘Calculated from experimentally determined rate constants

between 15 and 45o. dCalcuLated from experimentally determined rate constants between 165 and

2000. Values of 42.3 2 0.8 for AH* and 17.5 + 1.5 for AS* were obtained from calculations

using the gas phase rate constants determined by A. U. Blaekham and N. L. Eatough, 2. &. Chew

K., 84, 2922 (1962). %lculated from experimentally determined rate constants between 145

and 175”. f Calculated from experimentally determined rate constants between 120 and 150°.

gcalculated from experimentally determined rate constants between 105 and 135O. hCafculated

from experimentally determined rate constants between 120 and 147”.

No.46 4631

Results of our studies on the decomposition of Ic show a similar rate acceleration to

result from the substitution of a second pair of cyclopropyl groups into the AIBN molecule.

If the driving force for the decomposition originates from concerted ring cleavage, then the

maximum additional rate acceleration between Ib and Xc would be statistical in origin (a factor

of two for a mechanism involving cleavage of one ring, V, or four if two rings were simultaneously

involved, VI) . The difference between Ib and Ic, 9.3, is greater than the maximum statistical

consideration, 4, and makes concerted ring cleavage as the sole provider of the driving force

seem unlikely. _

F’N-N--~~ J ‘4 *-- 6- --C--N = N---C__ ---

c!N %

CN LN r!N

V VI

From the decomposition of Ic in the presence of a slight excess of hexaphenylethane we are

able to account for 78% of the material decomposed as reaction products of the dicyclopropyl-

cyanomethyl radical in which both cyclopropyl rings are maintained intact. The products, 2,2-

dicyclopropyl-3,3,3_triphenylpropionitrile (VII, mp 139-140’) and tetracyclopropylsuccinonitrile

(VIII, mp 78-79’), were identified from their elemental analyses and infrared and nmr spectra.

&Y-N=N-~4 V

LN LN

hexaphenylethaneO

Ic VII (53%) VIII (25%)

To the extent that decompositions of azonitriles proceed through transition states re-

sembling radical products, the activation energies in Table 1 reflect the stabilizing inter-

action of cyclopropyl substftuents with adjacent radical centers. It can be seen that the

effect of replacing each pair of methyl groups with a pair of cyclopropyl groups in the series

Ia, Ib, and Ic is roughly additive in free energy. The first pair of cyclopropyl substituents

lowers AG* by 2.3 kcal/mole and the second by 1.5 kcal/mole.

The nature of these interactions might, however, involve charge polarization of two types

which uould remove electron density from the central carbon atom of the incipient radical in

the highly polarizable transition state. The first type, represented by resonance structures

IX and IL, is of a type expected when a bond joining two atoms of different electronegativity

4632

is stretched or when an electronegative atom abstracts hydrogen from a C-H bond (2).

of charge polarization is a possible factor in all of the kinetic studies from which

have been drawn relative to the stability of cyclopropylcarbinyl radicals. In fact,

cases where polarization might be expected not to be large, accelerations of radical

No.46

lhi6 8Ort

argument8

in creveral

formation

attributable to interactions of cyclopropyl groups are very minor (la), or essentially undetect-

able (5). The decomposition8 of type II azo compounds, thought to proceed by simultaneous

cleavage of both C-N bonds (6). presents a case in which sysssetry considerations make it seem

unlikely that transition state polarieation is much greater than ground state polarization.

Contributions from structures IX and X place negative charge on adjacent nitrogen atoms and

would be mutually delimiting.

+- -+ [R-N-N-R +--a RN-N-R +----) R-N==N R]*

IX X

The second type of charge polarization, which might be expected to be nresent in

decomposition of aZObi8nitrile8, is illustrated by resonance structures XIII and XIV. These

suggest a possible resemblance of the transition state to the very stable cyclopropyl carbonium

ion (3,7). In order to determine the importance of this type of polarization as a possible

contributor to accelerations seen for Ib and Ic, we studied the rates of decomposition for a

series of aso compounds of type II, lacking the cyan0 function. Compounds IIa (8); IIb,

+

N-

XI XII XIII XIV

bp 43-45’ (hum); IIc, bp 80-81’ (0.25~~s); IId, mp 17-18.5’; and IIe, mp 45.5-46.5’, were

prepared by oxidative coupling of the corresponding amines with iodine pentafluoride (8).

Complete synthetic procedures will be reported in a later publication. In all cases satis-

factory elemental analyses and nmr and infrared spectra were obtained. The decomposition rates

were determined in the temperature range 105-200’ in diphenyl ether with a small amount of

added isoquinoline as a stabilizer against acid-catalyzed decomposition. Again the addition

of each pair of cyclopropyl substituents produces an approximately additive change in the free

No.P6 4633

energy of activation. The AAG’ for progression from IIa to III is 2.9 kcsl/mole; III to IIc,

2.1 kcal/mole; and IIc to IId, 1.5 kcal/mole. The deviation from strict linearity in AG* could

be the result of a saturation effect or, more likely, an increasing steric inhibition of reson-

ance as the number of cyclopropyl groups is increased. The close similarity in relative rates

and free energy of activation in compounds IIc and IIe provides evidence agafnst a contention

that the increase in rate observed in the series IIa, IIb. IIc, and IId is due to an increase

in steric repulsion in the ground state that is relieved in the transition state.

Consideration of the data in Table 1 provides an argument against reasoning that

acceleration is provided for by relief of ping strain, and the similarity seen in &W* values

in the two series of Table 1 indicates that the acceleration of decomposition rate seen on

substituting cyclopropyl for methyl in .an azornethane does not depend on developing carbonium

ion character at the central carbon in the transition state, as in XIII or XIV, but reflects

a stabilizing interaction between the cyclopropyl substituent and a developing radical center.

Decomposition of l,l,l,l’.l’,l’-hexacyclopropylazomethsne (IId) in 1,4-cyclohexadiene

yields more than 80% of l,l-dicyclopropyl-1-butene and a small amount (2%) of hexacyclopropyl-

l thane . The latter compound was identified by comparison with an authentic sample prepared

in 402 yield from photolysis of a sample of III at 0'.

lexacyclopropylethane decomposes at 295’ in decalin with a first-order rate constant of

1.31 x 10 -3 -1

set . This corresponds to a free energy of-activation of 41.5 kcal/mole. Using

a value of 13 eu for MS for the dissociation of hexacyclopropylethane into two tricyclo-

propylmethyl radicals (A8 for III, Table l), and assuming that (a) solvation effects are un-

important and (b) that the energy of activation for recombination of two tricyclopropylmethyl

radicals is small, we arrive at a rough estimate of 45 kcal/mole for the bond dissociation

energy of hexacyclopropylethane. The low value of this estimate suggests that decomposition

does not proceed by simple cleavage of a cyclopropane ring (9). It is, however, impossible

to rule out ring cleavage concerted with central C-C bond cleavage.

The estimate for the bond dissociation energy of hexscyclopropytethane, 45 kcal/mole,

while larger than that of hexaphenylethane, 15 kcal/mole, is considerably less than the

estimated value of 67.5 kcal/moIe for 2,2,3,3_tetramethylbutane (10). Ihese results are

consistent with the postulate that tricyclopropylmethyl radicals are resonance stabilized,

although the differences in bond dissociation energy may, in part, be attributed to differences

4634 No.46

in steric interactions.

Acknowledgement. - The work was supported in part by a grant from the U. S. Public

Health Service (GM-12296) and in part by a grant from the U. S. Army Research Office (Durham).

1.

2.

3.

4.

5.

6.

7.

a.

9.

References

(a) D. C. Neckers, A. P. Schaap, and J. Hardy, 2. &. 9. &., 88, 1265 (1966);

(b) E. S. Huyser and D. T. Wang, 2. w. _Chem., 29, 2720 (1964); z) E. S. Huyser

and J. D. Taliaferro, E., 28, 3442 (1963); (dF C. Walling and P. S. Fredricks, 2.

Am. Chem. &., 84, 3326 (1962); -- (e) H. Hart and D. P. Wjman, u., g, 4891 (1959);

(f) H. Hart andT. A. Cipriani, s., 84, 3697 (1962); (g) C. G. Ovzberger and

A. Lebovits, g., 76, 2722 (1954); (hr C. G. Overberger and M. B. Berenbaum, w.,

73, 2618 (1951). - - W. A. Pryor, "Free Radicals," McGraw-Hill Book Co., Inc., New York, N. Y. 1966, p. 170.

H. Hart and J. M. Sandri, 2. &. Chem. E., 8l, 326 (1959); H. Hart and P. S. Law, -

s., 86, 1957 (1964). -

C. G. Overberger, M. Tobkes, and A. Zweig, J. 9. m., 28, 620 (1963). -

D. C. Neckers and A. P. Schaap, w., 32, 22 (1967). -

S. Seltzer, 2. &. a. &., 83, 2625 (1961); S. Seltzer, w., 85, 14 (1963);

S. G. Cohen and C. H. Wang, ibi;rr, 77, 3628 (1955); C. G. Overbergec J-P. Anselme,

and J. G. Lombardino, "Organic Compounds with Nitrogen-Nitrogen Bonds," The Ronald

Press Co., New York, N. Y. 1966, p. 32.

R. Breslow in "Molecular Rearrangements," vol. 1, P. de Mayo, Ed., Interscience Publishers,

Inc., 1963, Chapter 4.

T. E. Stevens, 2. a. 9., 26, 2531 (1961). -

D. W. Setser and B. S. Rabinovitch, 2. Am. Chem. h., 86, 564 (1964). -

10. S. W. Benson, 2. Chem. Ed., 42, 502 (1965). -